Truth Beyond the ‘Trigger Puller’: Moral Accountability, Transitional (In)Justice and the Limitations of Legal Truth


 This article critically examines how existing truth recovery processes have, and how proposed truth recovery processes might, address the issue of shoot-to-kill in Northern Ireland. In contrast to ongoing legacy case prosecutions of British Army veterans, it argues that truth recovery processes should adopt a maximalist conceptualization of truth and responsibility. This, it is argued, necessitates differentiating between legal truth and structural truth, and between legal, political and moral responsibility. It is further submitted that such an approach would reflect the importance of identifying and understanding patterns (looking at similar cases collectively) and context (looking at the socio-political-legal backdrop to them) when establishing the ‘broad circumstances’ around disputed killings. This approach would allow truth recovery to advance beyond low-level ‘trigger pullers’ to ensure ‘moral accountability’ for the complicity of other actors within the state apparatus.


INTRODUCTION
Truth is one of the key concepts found within transitional justice (TJ) literature, yet at the same time it defies any agreed definition. 1 There is, for instance, significant debate about whether truth is a process or a product and/or whether it is an end in itself or a means to another end. 2 In any society attempting to 'deal with the past' discomfiting questions naturally arise as to what truth should contain, why truth is necessary and how 'truth-seeking' should be done. 3 Northern Ireland (NI) is no exception to this. More than 20 years after the Good Friday Agreement (GFA) peace accord, a 'joined-up' cases where lethal force was used in mistaken self-defence within the ambit of domestic criminal law, where lethal force was used in a public order context and, more controversially and most relevant to this article, where lethal force was used by specialist military and/or police units to eliminate suspected insurgents in preference to arresting them. 12 An expansive body of case law on the procedural limb of the Article 2 European Convention on Human Rights (ECHR) right to life subsequently emerged out of the NI context, 13 eventually shifting the focus of examination off issues like proportionality and the actual use of force and onto the preparation for and the investigation of the use of lethal force. 14 Thanks to this body of jurisprudence, the procedural limb of Article 2 now demands that the 'broad circumstances' surrounding deaths caused by state agents be established. 15 While a strong body of scholarship has already discussed proposals to 'deal with the past' in NI, victims' legal campaigns for truth and, more latterly, the prosecution of military veterans, 16 this article bridges these strands of scholarship by more specifically critiquing how the structural nature of shoot-to-kill might be firstly conceptualized, and secondly addressed, by truth recovery. Cognisant of the wider debate about 'truth versus justice,' 17 and accepting that the right to know what happened is not inextricably linked to prosecutions, 18 it uses recent legacy case prosecutions as an entry point to discuss the 'broad circumstances' of shoot-to-kill in NI.
Echoing calls for TJ to move -or perhaps better put, keep moving -'beyond legalism,' 19 this article argues that truth recovery must capture legal responsibility, moral responsibility and political responsibility, and produce structural truth as well as legal truth. This imperative is of growing importance as TJ continues expanding beyond the paradigmatic transition to address new structural harms that cannot be truly 'dealt with' through a narrow legalistic approach. 20 The article is premised on the central argument that the 'broad circumstances' of structural harms, be that shoot-tokill, colonial injustices or historical institutional abuses, cannot be established through restrictively focusing on the direct legal responsibility of a transgressor on the ground without giving due consideration to those who hold moral and/or political responsibility through sponsorship, through enablership or through directorship. Nor can they be established by reaching a narrow legalistic ruling in any given individual case that fails to identify relevant patterns across similar cases and to locate such patterns within the prevailing socio-political-legal climate of the time.
Building from that premise, the article argues for a fuller concept of truth capable of satisfying the victims' right to truth and the newly emergent societal 'right to know,' 21 and of guaranteeing non-recurrence through 'learning lessons' about past abuses. Such a truth, it is argued, is necessary to ensure 'moral accountability' for structural harms -something that cannot be dispensed through punishing low-level 'trigger pullers' in the courts. 22 The prosecution of the 'trigger puller' often reduces structural and cultural patterns of violence to the transgressions of a few 'bad apples,' 23 overlooking how these same transgressions rely on wider structures, resourcing and a supportive political context. 24 Without acknowledging this, truth recovery can easily degenerate into 'transitional injustice' by scapegoating a few of the lowest level perpetrators. 25 The article opens by theoretically examining the concepts of truth and responsibility, before moving enquiry beyond the 'trigger puller' by discussing the politicallegal backdrop to the conflict in NI. Attention then turns to existing truth recovery efforts, critically examining how and why they have thus far fallen short of structural truth. The article concludes with some observations on how proposed mechanisms might plug existing truth recovery gaps by leveraging the importance of both patterns and context when determining the 'broad circumstances' of shoot-to-kill. Before all that, however, it is necessary to briefly sketch out the current state of play in NI.

'DEALING WITH THE PAST': THE CURRENT STATE OF PLAY
The most recent set of proposals for 'dealing with the past' -the 2014 Stormont House Agreement (SHA) -supports the joined-up approach: an investigatory police body to (re)examine conflict-related cases (the Historical Investigations Unit (HIU)), an information retrieval body to provide information to victims' families (the Independent Commission on Information Retrieval (ICIR)) and an oral history archive for people to record their personal experiences of the conflict (the Oral History Archive (OHA)) would each perform their separate functions while simultaneously feeding into the Implementation and Reconciliation Group (IRG) responsible for conducting a thematic overview of the past. 26 However, as with previous proposals, the SHA has not yet made it to the implementation stage because of wider legacy politicking.
Part of the SHA fallout centres on political and legal disagreement over prospective prosecutions against former security force personnel. Whether the proposed HIU keeps the 'justice option' open for victims of state violence or whether it represents an alleged imbalance against former security force personnel who were cleared in previous investigations remains a key area of contention. 27 However, the likelihood of further legacy case prosecutions, and indeed the future of the SHA itself, are uncertain. In March 2020, in what was a significant unilateral departure from the SHA, the UK government controversially proposed an end to the reinvestigation of legacy cases without 'a realistic prospect of a prosecution as a result of new compelling evidence.' Instead, it would prioritize 'providing as much information as possible to families about what happened to their loved ones.' 28 Then, in May 2021, the UK government signalled its intention to bring forward legacy legislation within the coming year; 29 whether this will be the belated implementation of the SHA or a renewed attempt to prevent ongoing and future prosecutions against former British soldiers remains unclear.
In Truth Beyond the 'Trigger Puller' 5 TYPOLOGIES OF TRUTH AND RESPONSIBILITY Although by no means immune from arcane disagreements over the concept of 'truth,' 32 the TJ literature contains fruitful discussion on the 'right to truth,' whether truth-telling represents a form of accountability and how truth might be a form of reparation. 33 The case still remains, however, that there are different understandings of truth, varying degrees of truth and different versions of the truth, 34 making it plural and contested rather than singular and certain. 35 However, as Ignatieff argues, the aim of any truth recovery process should be to reduce the amount of 'permissible lies' in dominant discourses 36 -an argument that in itself reveals the power dynamics behind how discourse ordains what is or is not the 'truth.' 37 From this perspective, truth recovery might be seen as determining what is true through the process of eliminating that which can be shown to be untrue -in other words, that deemed impermissible lies.
Truth recovery in the South African context represents an instructive example. The report of the Truth and Reconciliation Commission (TRC) offered four paradigms of truth: factual/forensic truth, personal/narrative truth, social/dialogue truth and healing/restorative truth. 38 Without discounting the importance of the other two conceptualizations, this article concentrates on factual/forensic truth and personal/narrative truth. It does so for two reasons; because they are the most likely to narrow the range of 'permissible lies' about shoot-to-kill, and because they are the variants that are either explicitly or implicitly discernible in truth recovery efforts to date. It is therefore necessary to unpack these two typologies and clarify how they are conceptualized and further developed in the present study.

Legal Truth
Factual/forensic truth is derived using legal or scientific processes to establish an overview of what happened, whom it happened to, when it happened and where it happened. 39 Within a legal process or legal forum, 'truth' is essentially understood to be factual information retrieved through evidential procedures and evidential thresholds that answers basic questions of who, what, when and where. 40  questions are answered it provides what Robert Summers calls a 'formal legal truth' of true or false, innocent or guilty and permitted or prohibited. 41 The appeal of this type of truth is the apparent strength of the processes used to determine it. 42 As a matter of clarity, this article uses the term legal truth to denote forensic/ factual truth that has been gleaned through a legal forum or legalistic process. Key to this conceptualization is the fact that it has the strength of being obtained through a formal legal process or, at the very least, of being accepted in an official legal forum.
For all its procedural and evidential strengths, legal truth must be critically evaluated in terms of its ability to reduce 'permissible lies.' On one level it can do this. Take for example the inquest into the death of Derry teenager Manus Deery who was shot dead by the British Army in May 1972. The legal truth as established therein was that, contrary to official reports at the time, Deery was not armed when he was shot nor was he in the IRA. On the back of this legal truth the Ministry of Defence (MoD) was forced to concede that the killing was 'unjustified.' 43 Through the process of eliminating the original British Army claims as impermissible lies, the official account of an IRA gunman being justifiably shot dead was deemed untrue.
On the other hand, there are fundamental limitations to the degree of 'permissible lies' that legal truth can reduce. Its principal weakness is that it obscures the relevance of macro-level factors in favour of focusing on barest fact in any given individual case. The shortcoming here is that, as Osiel argues, 44 attention is being paid to what is ultimately only a very small piece of a much larger jigsaw. It discounts wider contextual issues and structural processes that go beyond whether X did or did not do Y to Z. 45 Likewise, in seeking only to determine if something is legal or illegal it reduces the issue to one of legal responsibility. This obscures the concomitant moral and political responsibility for systematic violence by reducing it to a binary judgement that either incriminates or absolves on a narrow legal basis. 46 Hayner thus argues that criminal guilt, in the guise of a 'formal legal truth' as determined by a trial, and responsibility for wrongdoing, as adjudicated by a truth recovery process, are not the same thing. 47 Whereas legal responsibility narrows the issue to one of criminal liability derived through causality, moral responsibility is instead premised on the wider concept of answerability. 48 By avoiding the simple causal basis adopted by criminal liability, moral responsibility inculpates those with knowledge of and influence over certain acts from which they are more causally removed in any immediate sense. 49 Given their ability to set the parameters of what society deems to be legally and morally permissible, the things that statesmen and politicians say and/or do have real world consequences. 50 In contexts of political violence and armed conflict, as Oren Gross notes, public officials may champion extra-legality when demanding a tougher state response and then subsequently provide political cover for any such response. While their tough talk benefits from the protective ambiguities of a 'legal black hole,' 51 the actions of the 'trigger puller' do not. Even if these actors do not bear legal responsibility, they do bear political responsibility.
In the case under consideration, this means that while a criminal trial would establish a 'formal legal truth' in relation to any wrongdoing by the state agent who pulled the trigger, it would not examine the moral or political responsibility of others beyond this.

Structural Truth
Rather than asking the rudimentary questions of who, what, when and where, narrative truth draws out how systematic violence came to happen and why it was tolerated. 52 Instead of individuating responsibility, it takes a broader approach of scrutinizing those who often escape formal censure via legal processes. 53 It looks beyond the 'basic responsibility' of those most often held causally and/or legally responsible (i.e., the person who pulls the trigger) to consider the 'consequential responsibility' of those whose acts and/or omissions had particular moral consequences for structural violence. 54 By doing so, it interrogates how cultures of secrecy and impunity benefitted perpetrators on the ground, thus opening up to scrutiny those who ordered, concealed and supported state violence through word or deed. 55 Where shoot-to-kill is concerned, this would extend consideration beyond the legal responsibility of the 'trigger puller' by including the political responsibility of sponsors who provided political encouragement and justification for the trigger to be pulled, and the moral responsibility of enablers who tolerated the trigger being pulled through silence and inaction and of directors who devised particular policies and practices for the trigger to be pulled. This approach reflects how truth recovery needs to be not only descriptive in terms of outlining 'what' happened but also explanatory by establishing 'how' and 'why' it happened. 56 For Ignatieff the imperative to establish 'how' and 'why' rests on key differences between 'factual truth' and 'moral truth,' and between narratives of the past that tell what happened and narratives that try to explain what happened. 57 Structural truth, as referred to in this article, represents a narrative truth that is capable of expanding on the factual basics of legal truth -the 'what' -by locating it within a framework that recognizes the political-military reality of the conflict in the North of Ireland -the 'how' and 'why.' It entails taking trends that can be pieced together through a series of legal truths and analysing them within what is known more widely about the conflict itself. This has been labelled the 'social scientists' approach' to truth recovery, 58 and is discernible in existing studies examining when, where and how shoot-to-kill incidents took place. 59 Yet a structural truth that sufficiently addresses the 'broad circumstances' of shoot-to-kill will have to go beyond simply aggregating individual legal truths. As former UN Special Rapporteur Pablo De Greiff has argued in his criticism of the 'eventbased' approach taken thus far in NI, systematic violence cannot be approached as 'simply the aggregate of isolated events' but must also recognize the import of 'patterns, structures, institutions, organisations, chains of command and policies.' 60 Successfully attaining a structural truth, then, depends on the SHA mechanisms drawing the relevant lines of connectivity between the individual case, similar thematic cases and macro-level political attitudes and judicial approaches towards state violence at the time.

COUNTERINSURGENCY AS A 'WHOLE-OF-GOVERNMENT'
PHENOMENON The targeted removal of those (allegedly) engaged in political violence is invariably endorsed in some way at some level within some arm(s) of the state. 61 Aggressive counterinsurgency measures are, after all, often preceded by political and legal speak about toughening up on 'terrorism,' 62 making counterinsurgency a 'whole-of-government' strategy that combines the soft power of politics with the hard power of security. 63 Responsibility is thus a more complex affair than the basic legal responsibility of a low-level perfunctory found on the end of a trigger; it runs along a continuum with those bearing basic responsibility (i.e., the 'trigger puller') at one end and those bearing consequential responsibility (i.e., directors who gave orders, sponsors who provided political, moral or legal cover and enablers who turned a blind eye to or covered up obvious wrongdoing) at the other end. Shoot-to-kill in its proactive permutation was most notably, though not exclusively, used in NI as the state sought to contain IRA violence to an 'acceptable level' by the late 1980s. 64 The 'whole-of-government' approach during this containment phase merged a criminal enforcement strategy of modifying due process with a military strategy of selectively targeting those engaged in political violence. 65 Despite the sudden rise in the number of such killings, state officials denied that a shoot-to-kill policy was in operation. James Prior, the then NI Secretary of State, epitomized this denial when he said 'there is no shoot-to-kill policy, there never has been, and as far as I am concerned there never will be.' He further maintained, through a subsequent interpretive denial, that there was 'a policy of capturing and dealing with terrorists.' 66 Although denying a shoot-to-kill policy was in operation, the state argued that the 'special circumstances' in NI necessitated a relaxation of traditional legal standards on lethal force. 67 Legal scholars, though, have long argued that a counterinsurgency context did not grant armed state agents carte blanche over the use of lethal force. 68 While often engaging the rhetoric of 'war' when it was politically expedient to do so, the state avoided making any formal legal declaration of war in NI, choosing instead to use the ambit of 'public emergency' to implement counterinsurgency policies on the basis that political violence was a matter to be addressed through a combination of domestic criminal law and semi-permanent 'emergency' provisions. This rejected International Humanitarian Law (IHL) standards in favour of a model ill-suited to regulating the onground reality of a de facto low-intensity conflict. 69 While the level of armed activity fluctuated throughout the conflict in the North of Ireland, in the earliest phases the requisite threshold for IHL to apply had arguably been met. 70 Yet the problem goes beyond what stage of the conflict contested killings took place; when the IHL threshold may have been met in the earlier stages of the conflict most of those killed by the security forces were unarmed and not involved in the conflict (i.e., non-combatants under IHL), whereas even though the case for IHL applying by the late 1980s is weaker many of those killed then were in fact members of non-state armed groups. 71 In any case, it was not until after the GFA that the UK government finally ratified the Additional Protocols to the Geneva Convention that cover non-conventional armed conflict, 72 evidencing yet again its sensitivities around the conflict in NI being legally and politically recognized as an internal armed conflict. The UK government's present-day insistence on addressing legacy in accordance with 'rule of law' understandings of the conflict suggests that such sensitivities remain.
The security forces in NI were therefore bound by domestic criminal law provisions on the use of lethal force rather than IHL. Section 3 (1) of the Criminal Law Act (Northern Ireland) 1967 provides that: a person may use such force as is reasonable in the circumstances in the prevention of crime, or in effecting or assisting in the lawful arrest of offenders or suspected offenders or of persons unlawfully at large.
The question naturally arises as to whether a provision intended to regulate occasional reactive force in domestic contexts was ever going to be capable of regulating more frequent proactive force exercised in counterinsurgency operations. Ní Aoláin has argued that despite counterinsurgency being a context with which Section 3 of the Criminal Law Act 'was never designed to cope,' the state nonetheless tried to force a fit rather than admit that 'different rules are in operation.' 73 Although the problems posed by applying domestic provisions to a conflict context could in theory have been addressed through the ECtHR, strategic use of ex gratia payments following contested deaths precluded the issue, at least initially, from reaching this theatre. 74 Consequently, counterinsurgency strategy was scrutinized domestically where the security forces could be afforded greater latitude amidst political pleas of the 'special circumstances' argument. 75 As a result, few criminal trials followed the use of lethal force, and where they did, judges erred on the side of caution by being 'too subjective . . . to judging the actions of soldiers and police officers on the ground.' 76 Claims of victims' sudden suspicious movements, threatening body language and failure to heed warning soon became a common, and uncritically accepted, plank of self-defence arguments. 77 While assessing reasonableness and subjective mistaken belief is a problem central to any self-defence claim, it was in the NI context that they were leveraged to the advantage of the security forces under the 'whole-of-government' approach. This exposes a judicial reticence to constrain security policy when performing to a 'political audience' intent on a tougher counterinsurgency approach that deviated from the norms of domestic law. 78 This 'political audience' had already shown open political sponsorship for shoot-to-kill, with parliamentary debate following controversial deaths arising out of undercover security operations in Lurgan and Armagh hearing calls for suspected republican insurgents to be 'shot Truth Beyond the 'Trigger Puller' 11 like rats in the streets' and for the security forces to 'put the dogs down one way or another.' 79 Accordingly, as Ní Aoláin points out, judges created a legal landscape where it was permissible to use force in a 'fleeing felon' context even if the fleeing person was merely suspected of belonging to a non-state armed group rather than actually being armed or committing another offence; where soldiers must be judged on the mistaken facts they believed even if these mistakes were not objectively reasonable; and, not insignificantly, where any calculation of reasonableness should be limited to the immediate circumstances evident on the ground (i.e., the moment of opening fire) rather than taking into account operational pre-planning. 80 In light of these rulings, Dickson concludes that the domestic courts 'did little to ensure that the right to life was protected by law.' 81 This lacuna was only rectified when the families of three IRA members (Mairead Farrell, Sean Savage and Daniel McCann) shot dead by the Special Air Service (SAS) in Gibraltar took a case to the ECtHR. In finding that the failure to use intelligence to apprehend the deceased violated the Article 2 right to life, the court widened any examination to the 'surrounding circumstances' that included 'the planning and control of the actions under examination.' 82 Remarkably, though, in this case, and other similar Article 2 cases subsequently emerging from the North of Ireland, the court did not make any ruling on the actual use of force, concentrating instead on the broader context of planning and investigation. Nevertheless, in moving the focus off the armed agent on the ground, McCann certainly appears more conducive to a structural truth of shoot-to-kill than the domestic approach was.

THE LEGAL TRUTH OF SHOOT-TO-KILL
The existing piecemeal approach to truth recovery in NI has not produced a structural truth reflecting the 'broad circumstances' of shoot-to-kill, though it has provided helpful, if limited, legal truths. These legal truths can, however, be analysed collectively, with a view to further eliminating any impermissible lies. It is worth applying this approach to two of the mechanisms used so far to address shoot-to-kill; public inquiries and inquests.

Public Inquiries
The public inquiry is a narrowly focused inquisitorial mechanism designed to establish the facts of a particular event, to examine the structures and processes that led to the event, to identify the culpability of individuals and to symbolically provide some form of accountability through its findings. 83  used to deflect blame, close down the space for critical examination and reassert the legitimacy of the state and its way of 'seeing' the problem. 84 Nevertheless, a number of public inquiries have been established in NI, 85 with the Saville Inquiry into the shooting dead of 14 unarmed civil rights demonstrators by British soldiers from the Parachute Regiment on Bloody Sunday (30 January 1972) being the most notable example. Saville exonerated the dead and placed the blame on the soldiers involved, in stark contrast to the conclusions made in the original tribunal that had been led by Lord Widgery months after the killings. Notably, Saville used legal truth to dismiss Widgery's 'strong suspicion' 86 that nail bombers and gunmen were among the deceased as impermissible lies. Saville concluded that 'none [of the killed and wounded] was posing any threat of causing danger or serious injury,' 87 and that there were instances where 'soldiers shot at civilians who were trying to escape, attempting to surrender, or making efforts to tend to the dead or dying.' 88 Witness testimony and the disclosure of other evidence placed further significant legal truths into the public domain; several of the soldiers involved on the day admitted lying to the Widgery Tribunal and to military investigators at the time, 89 senior military figures had been pushing for a more confrontational and aggressive approach on the day and, perhaps most significantly, one senior military figure had advocated the shooting dead of selected 'ring leader' rioters. 90 While these significant legal truths provided the sum parts for a fuller structural truth, this was not reflected in Saville's report. Instead, it chose to confine responsibility to the basic responsibility of soldiers on the ground at the expense of a more robust critique of the consequential responsibility of others. The subsequent outworking of this is, as Mark McGovern has recently argued, that in being the only person sent to trial Soldier F is shouldering responsibility for the rest of the 'trigger pullers' involved, 91 for those who ordered British troops into the Bogside and for those who commanded them on the day. 92 Others have similarly criticized the 'few "bad apples" caught up in the heat of the moment' premise of the report for neglecting the consequential responsibility of the political and military chiefs who pushed for a more aggressive approach to be taken on the day and then instigated a campaign of disinformation and cover-up in its aftermath. 93 McGovern goes further in his criticism by arguing that any truth on Bloody Sunday must recognize the pervasive environment of impunity for state violence at the time, with the relevance of this being seen in the failure to hold anyone accountable for killings in Ballymurphy by the same regiment some months previously. 94 Even if Saville did remove the slur cast upon the dead by Widgery as impermissible lies, it did not produce a structural truth. This has not escaped the attention of the Bloody Sunday March Committee, who mounted a 'Jail Jackson' campaign calling for the prosecution of General Sir Mike Jackson who was second-in-command of the regiment on the day and who had been involved in the flawed investigation into the killings. 95

Inquests
Inquests in NI fall into two categories; outstanding cases where no inquest was held and cases where the Attorney General has ordered a new one due to inadequacies in the original inquest. Historically inquests in NI were excluded from examining the broader circumstances of death and returning a finding of unlawful killing by the state, placing obvious limitations on their usefulness in examining shoot-to-kill deaths. 96 However, extensive litigation by victims' families eventually forced change in the inquest system. This culminated in the case of Re McCaughey, relating to the shooting of IRA members Dessie Grew and Martin McCaughey by the SAS in October 1990, holding that an Article 2 compliant inquest must not only establish 'by what means' but also 'in what broad circumstances' the victims were killed. 97 The attention afforded here to the relevance of 'circumstances,' as had been done earlier by the ECtHR in McCann, is further acceptance of the need to more thoroughly examine all the salient events that led up to the pulling of the trigger.
Some useful legal truths have thus emerged in post-Re McCaughey inquests. The Pearse Jordan -an IRA member who was killed by the Royal Ulster Constabulary (RUC) in 1992 -inquest is an instructive case. This established that Jordan was unarmed and shot in the back and not when turning to face Sergeant A who had shot him; that Sergeant A had lied about previous shoot-to-kill incidents he had been involved in; and that the specialist police unit Sergeant A belonged to was trained to shoot-to-kill rather than shoot-to-disable with a view to apprehension. 98 The Grew/ McCaughey inquest proved similarly fruitful. It revealed that the officer commanding the SAS unit in question was linked to an 'extensive number' of other lethal force cases. 99 Moreover, Soldier K's insistence that the SAS did not have a shoot-to-kill policy was belied by the disclosure of MoD documents suggesting otherwise. 100 Most significant, however, was the admission that intelligence received a few days earlier had alerted the security forces to the planned IRA operation, 101 yet this was used to have the pair killed rather than arrested when they were unarmed.
When analysed together, the Jordan and Grew/McCaughey inquests reveal patterns of the same members of the security forces being involved in several shoot-tokill cases; these same members being trained to shoot-to-kill; and the elimination of suspects taking precedence over arrest. Given that neither case is anomalous compared to other recorded shoot-to-kill cases, there is little reason not to extend any subsequent analysis. If this is accepted, then the legal truth recovered through the piecemeal approach can at least dismiss previous claims that specific operations were not intended to eliminate suspects as impermissible lies. For all that, these aggregated legal truths in themselves do not actually look beyond the 'trigger pullers' on the ground.

CONTEXTUALIZING SHOOT-TO-KILL PATTERNS
The structural truth of shoot-to-kill cannot be found within aggregated legal truths that are divorced from the macro-level context. It is context that frames identifiable patterns with a meaning that can answer 'how' and 'why,' thus allowing the nuances of 'moral accountability,' as opposed to the simplistic binary thinking of legal accountability, to be addressed. Attempts to limit transparency around decisionmaking processes, policies and structures that influenced on-ground harms represent a case of 'transitional injustice' par excellence precisely because they restrict 'moral accountability' for structural harms to the 'trigger puller. ' 102 This has been relevant in the NI context where the state has imposed 'national security' limitations under the piecemeal approach. For example, then NI Secretary of State Theresa Villiers tried to prevent the Attorney General ordering a fresh inquest into the shooting of an IRA unit by the SAS in Loughgall because there is 'information, the disclosure of which may be against the interests of national security.' Whatever the reason cited, human rights group Committee on the Administration of Justice (CAJ) criticized the move precisely because it prevented any structural truth emerging: We don't know why this case has been singled out for the 'national security' veto -is it because the SAS were involved, or did a UK Minister give the green Members of the security forces have, for instance, spoken of later celebrating particular shoot-to-kill deaths 111 -this macabre triumphalism often being publicly echoed by politicians. 112 The significance of this is twofold; firstly, it again suggests that such operations were intended to eliminate rather than apprehend (botched arrest operations would hardly give cause for celebration), and secondly it evidences a level of political support for shoot-to-kill as had politicians been at all perturbed they would not have congratulated those involved.
Other accounts in the public domain provide an insight into the culture of impunity surrounding shoot-to-kill. Former members of the specialist Military Reaction Force (MRF) have claimed that security chiefs granted them a de facto shoot-to-kill licence to kill unarmed civilians without fear of prosecution in the early 1970s. 113 These claims are supported by the minutes of a 1972 meeting of political, policing and military officers that discussed the need to 'indemnify' security force members who shot dead civilians in communities 'harbouring gunmen and bombers.' 114 A previous memo from a senior military official to the then Attorney General encouraged him to 'do all within his power to protect the security forces from criminal proceedings in respect of actions on duty.' 115 Indeed, some accounts proffered by former soldiers highlight the disconnect between official discourse on the use of lethal force and the on-ground reality whereby soldiers adhered to, sometimes with open encouragement from their section commanders, a 'shoot first, questions later' approach. These accounts -written as exposés long before the current legacy debate -provide an insight into the acute awareness that 'trigger pullers' had of the political support and judicial sympathy they could rely on should lethal force rules be transgressed. 116 This creates a more complex picture of moral responsibility because the relevant question is not to ask whether any particular security force agent unjustifiably pulled a trigger but to ask if they did so with the peace of mind that political and military chiefs would shield them from legal accountability. The structural truth to be found within such accounts, then, supersedes the 'trigger puller' to implicate those directors, sponsors and enablers who helped to create, and then subsequently to sustain, the environment for the 'trigger puller' to knowingly operate with impunity.
Perhaps this point is best understood through Lord Justice Gibson's remarks in the trial of RUC officers arising from a 1982 shoot-to-kill case in County Armagh. During the trial Gibson not only commended those involved for their 'courage and determination' in sending the victims to 'the final court of justice' but also castigated the Director of Public Prosecutions (DPP) for bringing charges in the first place. 117 On the back of this, one may well ask whether Gibson bears some consequential responsibility for shoot-to-kill; the most critical reading would suggest that at the very least Gibson created a legal climate where shoot-to-kill became an acceptable component of the counterinsurgency containment strategy through word (i.e., sponsor), and at worst he actually facilitated it by shutting down the space for legal accountability through deed (i.e., enabler). A less critical interpretation would be that Gibson was simply implementing the rule of law at the time as was his job to do so. However, as David Dyzenhaus has argued in relation to judges in South Africa, this should not be grounds for automatically absolving him of moral responsibility. 118 Brice Dickson has similarly pointed out that whatever constraints were on judges like Gibson in the NI context, at the very least they could have expressed concern about the erosion of the right to life through aggressive counterinsurgency strategies. 119 Indeed, Gibson's comments can be usefully contrasted with those of Justice Horner at the recent Jordan inquest. Horner remarked that, whatever people may think, IRA membership does not: mean that such a person had in some way forfeited the right to be protected by the law . . . he does not become an outlaw who can be summarily executed whether by officers of the state or otherwise. 120 Yet Gibson's remarks are only one component of a complex picture of moral responsibility in this case given that the then Attorney General also held that it was not in the 'public interest' to proceed with prosecutions for perjury against RUC officers who had lied about the incident. 121 Politicians, too, may bear a moral responsibility beyond the political responsibility for sponsorship that has already been discussed. The killing of three IRA members by the SAS at Drumnakilly in August 1988 is the best illustration of this. Ten days before Drumnakilly the IRA had killed eight British soldiers in nearby Ballygawley, with then Prime Minister Margaret Thatcher subsequently convening emergency meetings to discuss recent IRA violence. What has since emerged about one particular meeting epitomizes debates over moral responsibility and 'moral accountability.' In a 2015 television interview former Unionist politician Ken Maginnis revealed that he had met with Thatcher after the Ballygawley attack and disclosed to her the identities of those he believed to be responsible. By his own admission, those he identified were shot dead at Drumnakilly days later. 122 This revelation exposes a complex chain of events reflecting the 'whole-of-government' backdrop to shoot-to-kill; an IRA attack sparks outrage, a political meeting is held to discuss a clampdown on IRA violence, those responsible are identified and then some days later they are shot dead in disputed circumstances.
The outrage -political reaction -disputed killing choreography that surrounded this case is also discernible in other disputed killings. For instance, cases involving the RUC at Armagh and Lurgan in 1982 can be traced back to previous attacks that also saw Unionist politicians pressing the Prime Minister to respond more aggressively to those allegedly involved. 123 This choreography suggests that responsibility extends well beyond those who pull the trigger on the ground, something that belatedly putting the 'trigger puller' in the dock, as legacy case prosecutions seek to do, fails to adequately capture. On that basis, it is difficult to see how 'moral accountability' can be dispensed without recognizing the complicity of those who created the context within which the 'trigger puller' operated.

CONCLUSION
Divorcing actions on the ground from what came before and after them limits what 'truth' is understood to be in transitioning societies and reduces the scope for 'learning lessons' from past abuses. Truth recovery, then, should orientate towards a structural truth that locates identifiable patterns within their relevant context, avoids selfimposed limitations on how truth and responsibility are understood and enjoys flexibility as to the potential (non-legal) sources available. This involves, as Ron Dudai puts it, 124 'recoding' all the information that we have about the past in order to 'unpack' the moral and political responsibility of those who were complicit through 'being indifferent and silent' and/or complicit through being 'active and demanding.' 125 If the 'broad circumstances,' to use the Article 2 speak, of state violence are to be truly understood then the contributory role of those enablers, sponsors and directors physically removed from the theatre of action that the 'trigger puller' operated within must be sufficiently scrutinized. This 'bigger picture' cannot be found through reductively seeking a legal truth via a criminal trial. The inadequacy of legal truth, however, extends well beyond shoot-to-kill cases in the North of Ireland. Indeed, as TJ mechanisms become increasingly drawn upon to address the legacy of various types of armed conflict, historical institutional abuse, colonialism and possibly now the Trump era in the US, 126 the inadequacy of legal truth in both delivering 'moral accountability' for and 'learning lessons' from longstanding structural harms only becomes more pronounced. The harms committed within the Magdalene Laundries in Ireland cannot be understood without acknowledging the morally toxic role that the state permitted religious institutions to play in everyday life, 127 just as an understanding of recent violence in the US cannot ignore political statements that instigated or excused it. This is not to say that legal truth does not have a role to play in TJ; it may indeed prevent impunity if it is used to punish an individual perpetrator. However, prosecuting the lowest level 'trigger puller' takes a reductionist view of accountability and impunity because it does not speak to the 'broad circumstances' of structural harms. If 'lessons learnt' are to prevent recurrence of, and impunity for, structural harms then this necessitates TJ with a structural truth recognizing the legal, political and societal particularities of the context within which structural harms occurred rather than 'transitional injustice' that chooses to ignore this when punishing low-level perpetrators.